search

facebookresearch / meshtalk

Code for MeshTalk: 3D Face Animation from Speech using Cross-Modality Disentanglement
Other
368 stars 56 forks source link

Does the pretrained model works for mandrain audio? #11

Closed lucasjinreal closed 2 years ago

lucasjinreal commented 2 years ago

Does the pretrained model works for mandrain audio?

alexanderrichard commented 2 years ago

It is trained on English language. We tried it on other languages and it produced reasonable results, though they were slightly worse than the results on english language. You can have a look at the supplemental video, there are some other languages in it.